Natural beauty used to mean artisanal soaps with salt crystals, or some avocado butter with essence of grapefruit; we eat it so it must be safe, right? Well, not entirely, and considering that natural extracts are composed of dozens if not hundreds of substances, a systematic approach to evaluating safety should be employed. Furthermore, product specifications and characterisation should be considered given that composition of natural extracts have shown to vary temporally and geographically and therefore may have safety impacts. This is further compounded by the fact that different plant parts of the same species have greatly divergent compositions. Finally, as discussed in “Are natural ingredients and products actually safer?”, the boom in natural beauty is partly driven by perceived safety of naturally derived ingredients and products. As will be demonstrated below, public perception of risk based on the natural origin of these ingredients is not sufficient to infer safety. As such, a robust and systematic approach would be needed to identify risk to human health of these types of ingredients and products.
History of use
Perhaps the most prominent reason as to why natural ingredients and products are perceived as safe by the public is history of use. The thesis is that since we have used it for centuries, and is natural, it must be safe. The history of use as a food or traditional medicine may provide valuable information on safety, but this is not always the case. In general, history of use includes information on duration, nature, and objective of use. For herbal products, two terms are distinguished: traditional use for which limited scientific documentation is available, and established use which is supported by scientific publications.
Most of the foods commonly eaten have not been investigated systematically with respect to their impact on consumer health. Generally, they are presumed safe unless a significant risk has been identified in humans. However, it should be emphasised that the absence of evidence of toxicity is not the same as evidence of the absence of toxicity. Without specific investigations, only acute and severe adverse effects are likely to be identified. To demonstrate the safety of a food requires specific intake information together with data on health impact. Although epidemiological observations, clinical and other case reports and/or experimental human studies may exist for a restricted number of foods, health information is lacking for most food plants (Schilter et al., 2003).
To further emphasise reasoning as to why history of use should not be the only source to infer safety, one can look at the examples of Rubia cordifolia and bamboo extracts. Rubia cordifolia is a climbing perennial herbal plant, which is widely distributed in China and India. Its root and rhizome (called Qiancao in China and Indian madder in India) is a well-known phytomedicine used to treat a number of ailments and is widely used in Ayurvedic and Traditional Chinese Medicine. In addition, it is a kind of pigment utilized as a food additive and a dye for wool or fibre. Furthermore, extracts from the root and rhizome have shown anti-inflammatory, anti-oxidant, and neuroprotective effects amongst others (Shan et al., 2016). However, some components of Rubia cordifolia root and rhizome extracts have shown to be genotoxins and/or carcinogenic. For instance, rubiadin, which is a constituent of the extract, was reported to display carcinogenic potential. In the outer medulla, cytoplasmic swelling with basophilic changes and karyomegaly were observed in male F344 rats fed with rubiadin for one week whereas 26-week oral administration of the component induced atypical tubules, putative pre-neoplastic lesions, and karyomegaly. The results indicated that rubiadin may be a potent carcinogenic ingredient that targeted the proximal tubule cells in the outer medulla (Inoue et al., 2009). In madder pigment, alizarin, purpurin and 1-hydroxyanthraquinone were found to have similar effects as ethidium bromide, a typical DNA intercalator. They exhibited potential genotoxicity by implanting into the DNA of Escherichia coli, blocking gene expression and inducing cell death (Li et al., 2013).
Regarding bamboo shoots, they have been used in food for centuries. In India, shoots are used as vegetables and pickle products. In traditional Chinese medicine, bamboo shoots are used to ease labour and the expulsion of the placenta by inducing uterine contractions. A poultice of the shoots is often used for cleaning wounds and healing infections. Bamboo shoot decoction taken with honey is used to treat respiratory disorders. However, to most people bamboo shoots are best known as food. Fresh, dried, or fermented bamboo shoots are used in numerous Asian recipes. Nutrient analysis on freshly emerging juvenile bamboo shoots has shown high contents of amino acids, proteins, carbohydrates, vitamins, and minerals, and a low content of fat. As bamboo shoots age, the dietary fibre and moisture start to increase while vitamin and mineral contents decrease (Panee, 2015). However, bamboo shoots contain potentially toxic compounds called cyanogenic glycosides i.e. Taxiphyllin, which break down upon disruption of plant cells to form hydrogen cyanide (Satya et al., 2010). Furthermore, it has been reported that an ethanolic extract of the tender shoots of Bambusa arundinacea (dose of 300 mg/kg body weight per day) changed the structural and functional integrity of the epididymis and reduced fertility of male rats. A 7-day treatment markedly reduced the count and motility of sperm (Manonayagi et al., 1989), and decreased the fertility index by 85 percent (Vanithakumari et al., 1989).
It is reported that bamboo shoot contains cyanide as high as 0.8% (Satya et al., 2010). However, cyanide content is reported to decrease substantially following harvesting. Furthermore, various indigenous methods of reducing acidity/bitterness from fresh bamboo shoots have been reported and some of them include chopping of tender shoots into small pieces, partial drying of fresh shoots, boiling in water/salt water and draining or keeping shoots in hot water for 10-15 min or in water for a week at ambient temperature, etc.
Studies have reported the optimum cooking conditions that resulted in 97% reduction of hydrogen cyanide (HCN) were 98-102 oC for 148-180 min (Ferreira et al., 1995). Subsequently, other authors have mentioned that removal of HCN can be done by steaming bamboo shoots (Tripathi, 1998). Bhargava et al. (1996) reported removal of this during cooking shoots by changing water several times or by pre-soaking for a long time and subsequent changing the 2% salt solution (Bhargava et al., 1996). Wongsakpairod (2000) reported superheated steam drying removes HCN from bamboo shoot as Taxiphyllin decomposes at around 116 oC.
As such, the case studies above indicate that history of use is not sufficient on its own for establishing safety, and further highlight the importance of accurate ingredient characterisation in order to establish safety in a robust manner.
Natural ingredient/product characterisation
Plant oils and extracts differ from synthesized chemicals in their origin. Synthesized chemicals are based on chemical reactions whereas oils and extracts are manufactured by physically processing material of biological origin. The significant difference between oils/extracts and synthetic chemicals is the composition or specification. Synthetic chemicals can be produced in standardized processes and resulting in repeatable purities. The composition of an extract/oil, however depends on the material of biological origin and the manufacturing process(es). Therefore extracts/oils have a larger variation in the qualitative and quantitative composition compared to synthetic chemicals.
It is therefore essential that the material should be well identified and characterised given the variation seen in natural extracts. This is particularly pertinent since a compositional approach may be pivotal in establishing safety of a natural extract. There are three aspects to the production of a consistent material: identification of starting material, manufacturing process, and standardisation of final product. The extent of information necessary for an appropriate characterisation of the plant material should be decided on a case-by-case basis, taking into account all available relevant knowledge. The history of use of the raw material, the history of food application, the manufacturing process, the expected exposure to the new product and the anticipated levels of toxicants in the finished product as consumed are key information requirements to prioritise the type and extent of data necessary. The biological properties of botanical preparations should be reproducible from batch to batch and, therefore, product standardisation is necessary.
Toxicological approaches to establishing safety
The safety of natural ingredients should be evaluated as any other ingredient; safety should be established using traditional toxicological approaches where each endpoint is systematically investigated and a point of departure identified below which systemic toxicity is not expected after accounting for safety factors. Furthermore, local toxicity potential should also be investigated through validated tests. However, given the nature of natural extracts (unknown and variable composition), traditional approaches are often not suitable. This is further compounded by the animal testing ban for cosmetic ingredients. As such, other approaches are needed in a weight of evidence.
Systemic toxicity
For systemic toxicity, often animal studies on the extract are not available. As such, safety could be inferred by investigating the major components of the extract. However, this would initially require having the composition of the extract, which is often not the case. This is further exacerbated by the fact that natural extracts from the same species have shown to be dependent temporally and geographically at time of harvesting. For instance, extracts from C. athaeoides L. leaves collected in winter and spring seasons had a shown increased fungicide activity compared to summer and autumn (Hrichi et al., 2020). The authors further identified that polyphenolic compounds were predominant in extracts from leaves collected in winter, whereas pigments were predominant in the spring collections, further indicating that potential changes in biological activity result from seasonal changes in composition. Similarly, extracts from the same species have shown geographically dependent variation. As an example, Cyclocarya paliurus has been widely used as an ingredient in functional foods in China. Authors have identified that its in vitro and in vivo antioxidant activities varied between the Wufeng and Jinzhongshan regions (Liu et al., 2018). More specifically, the authors identified a difference of up to 6-fold in in vitro antioxidant assays, and that the accumulation or responsible antioxidant phenolics was driven by sun-hours and temperature. There are numerous other examples indicating temporal and geographical variations in natural extracts. Nevertheless, the above examples highlight that if compositional approach is to be taken in light of lack of animal data for systemic toxicity, special consideration would have to be given to potential impacts from seasonal and geographical variation.
In silico toxicological approaches could also prove beneficial in assessing risk if the natural ingredient has no reliable toxicological data. The threshold of toxicological concern (TTC) is a concept that refers to the establishment of a human exposure threshold value for all chemicals, below which there would be no appreciable risk to health (Kroes and Kozianowski, 2002). It was developed in the field of food and food additives in order to limit further toxicological testing for chemicals that are only present at very low levels. The US Food and Drug Administration (FDA) accepts a dietary intake of 1.5 μg/kg bw/d as sufficiently negligible as to not present a public health concern. The TTC principle is used by the European Medicines Agency (EMA) to assess genotoxic impurities in pharmaceuticals. It was moreover used by the former EC Scientific Committee on Food and is now used by the European Food Safety Authority to evaluate flavouring substances. It has been endorsed by the WHO and by the EU Scientific Committee on Toxicology, Ecotoxicology and the Environment. Authors have also previously shown that the TTC approach can be used for the safety evaluation of cosmetics and that it provides a useful additional tool for risk assessment of ingredients and impurities of known chemical structure in the absence of chemical-specific systemic toxicity data (Kroes et al., 2007). Recently, authors have suggested a TTC for natural extracts higher than the one for Cramer III substances (1.5 μg/kg bw/d) (Kawamoto et al., 2019). However, these TTC values have not been widely used yet. As such, if the compositional data indicates lack of genotoxic substance for the natural extract, the TTC approach of assuming Cramer III thresholds can be argued to being protective for human health, provided that genotoxicity can be excluded.
Local toxicity
For local toxicity, there are a number of validated in vitro tests available that can identify hazard for the actual natural extract. Furthermore, as with systemic toxicity, the compositional approach may be taken to identify risk for local effects. In silico approaches are also possible in the case of skin sensitisation. This is an endpoint of significant consideration for cosmetics in particular as they are applied dermally in general. In addition, it is also a considerable endpoint of interest for natural extracts given that fresh plant material typically contains 0-2% essential oil, which in turn is composed of many terpenes, terpenoids, and phenylpropanoids – many of which are known skin sensitizers.
For skin sensitizers it has been suggested that the TTC concept can be used to identify a dermal sensitization threshold (DST). If no toxicity data are available, this concept can be used to calculate safe concentrations for cosmetic raw materials with respect to the skin sensitization endpoint. The DST concept is based on the assumption that a general threshold for sensitization can be defined without testing the appropriate substance below which no risk to develop an allergy exists. The DST was originally developed based on a Local Lymph Node Assay (LLNA) dataset and applied to chemicals that were not considered to be directly reactive to skin proteins, and unlikely to initiate the first mechanistic steps leading to the induction of sensitization (Safford, 2008). A DST of 1.64 μg/cm² was derived for rinse-off cosmetic products (shampoo) and 0.55 µg/cm² for leave-on cosmetic products (deodorant). Therefore, the use of a DST approach would preclude the need for sensitization testing of ingredients where dermal exposure is sufficiently low.
Conclusions
Natural does not necessarily always mean safer. Firstly, there are many examples where substances and extracts of natural origin are more toxic than synthesized substances. As such, safety cannot be established solely on origin of ingredient and history of use. Instead, a robust and systematic approach is warranted when evaluating these complex ingredients and products. Furthermore, given the nature of natural ingredients (complex mixtures of substances) and the impact of preparation, geography, and time of harvest, robust characterisation of the ingredient or product is needed. Indeed, comprehensive compositional data could be used to substantiate safety provided that reliable toxicological data are available for major components. Such an approach has already been employed by the SCCS when evaluating tea tree oil (SCCP/1155/08). Furthermore, there are a number of validated in vitro assays able to identify hazard for local effects using the actual extracts. When data are lacking, a threshold of toxicity approach may also be employed. Finally, history of use may be used in a weight of evidence approach to further substantiate safety, but should not be the sole factor. Ultimately, a significant amount of data is required to establish risk to human health in a systematic manner, and natural origin does not necessarily indicate safety.
Owing to expertise in a number of relevant fields. kriTOX can help you define the best strategy for your products. Please click here should you need more information about cosmetic ingredient services.
Published by Boris Krivoshiev, PhD – Founder and Principal Toxicologist at kriTOX
References
Bhargava, A. et al. (1996) ‘Bamboo parts and seeds for additional source of nutrition’, Journal of food science and technology (Mysore), 33(2), pp. 145–146.
Ferreira, V. L. P., Yotsuyanagi, K. and Carvalho, C. R. L. (1995) ‘Elimination of cyanogenic compounds from bamboo shoots Dendrocalamus giganteus Munro’, Tropical Science (United Kingdom).
Hrichi, S. et al. (2020) ‘Effect of seasonal variation on the chemical composition and antioxidant and antifungal activities of Convolvulus althaeoides L. leaf extracts’, Arabian Journal of Chemistry, 13(6), pp. 5651–5668.
Inoue, K. et al. (2009) ‘Carcinogenic potential of alizarin and rubiadin, components of madder color, in a rat medium‐term multi‐organ bioassay’, Cancer science, 100(12), pp. 2261–2267.
Kawamoto, T. et al. (2019) ‘Threshold of Toxicological Concern (TTC) for Botanical Extracts (Botanical-TTC) derived from a meta-analysis of repeated-dose toxicity studies’, Toxicology Letters, 316(July), pp. 1–9. doi: 10.1016/j.toxlet.2019.08.006.
Kroes, R. et al. (2007) ‘Application of the threshold of toxicological concern (TTC) to the safety evaluation of cosmetic ingredients’, Food and Chemical Toxicology, 45(12), pp. 2533–2562. doi: 10.1016/j.fct.2007.06.021.
Kroes, R. and Kozianowski, G. (2002) ‘Threshold of toxicological concern (TTC) in food safety assessment’, Toxicology Letters, 127(1–3), pp. 43–46. doi: 10.1016/S0378-4274(01)00481-7.
Li, C. C. et al. (2013) ‘Study on potential toxicity of madder pigment’, Guangdong Agric. Sci, 40, pp. 35–36.
Liu, Y. et al. (2018) ‘Geographic variation in the chemical composition and antioxidant properties of phenolic compounds from Cyclocarya paliurus (Batal) Iljinskaja leaves’, Molecules, 23(10), p. 2440.
Manonayagi, S. et al. (1989) ‘Effects of bamboo buds: structural and functional changes in the epididymis of rats’, Journal of ethnopharmacology, 25(2), pp. 201–212.
Panee, J. (2015) ‘Potential Medicinal Application and Toxicity Evaluation of Extracts from Bamboo Plants.’, Journal of medicinal plant research, 9(23), pp. 681–692. Available at: http://www.ncbi.nlm.nih.gov/pubmed/26617977%0Ahttp://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC4659479.
Safford, R. J. (2008) ‘The Dermal Sensitisation Threshold-A TTC approach for allergic contact dermatitis’, Regulatory Toxicology and Pharmacology. doi: 10.1016/j.yrtph.2008.02.010.
Satya, S. et al. (2010) ‘Bamboo shoot processing: food quality and safety aspect (a review)’, Trends in Food Science and Technology, 21(4), pp. 181–189. doi: 10.1016/j.tifs.2009.11.002.
Schilter, B. et al. (2003) ‘Guidance for the safety assessment of botanicals and botanical preparations for use in food and food supplements’, Food and Chemical Toxicology, 41(12), pp. 1625–1649. doi: 10.1016/S0278-6915(03)00221-7.
Shan, M. et al. (2016) ‘A review of the botany, phytochemistry, pharmacology and toxicology of rubiae radix et rhizoma’, Molecules, 21(12), p. 1747.
Tripathi, Y. C. (1998) ‘Food and nutrition potential of bamboo’, MFP News, 8(1), pp. 10–11.